Skip to main content

CellValue Model (SRM)

The CellValue model is used to store SRM-related observational data.

Underlying Schema that models cell values of an SRM (inspired by EAV)​

Apache Ignite​

CREATE MEMORY TABLE SRM.CELLVALUE(
    -- Apache Ignite key (internal)
    _KEY OTHER INVISIBLE  NOT NULL, 
    -- Apache Ignite value (internal)
    _VAL OTHER INVISIBLE ,
    -- LSL Script execution ID
    EXECUTIONID VARCHAR NOT NULL,
    -- LSL Action ID
    ACTIONID VARCHAR NOT NULL,
    -- Abstraction container ID
    ABSTRACTIONID VARCHAR NOT NULL,
    -- Arena Execution ID
    ARENAID VARCHAR NOT NULL,
    -- Sequence Sheet ID (typically name of the test)
    SHEETID VARCHAR NOT NULL,
    -- ID of the code as it appears in the code index
    SYSTEMID VARCHAR NOT NULL,
    -- A variant of the code depicted by the SYSTEMID above (e.g., mutant code)
    VARIANTID VARCHAR NOT NULL,
    -- ID of a particular adapter for the code depicted by the SYSTEMID above as generated part of the adaptation process
    ADAPTERID VARCHAR NOT NULL,
    -- Sequence coordinate X (i.e., part of a statement, >= 0)
    X INT NOT NULL,
    -- Sequence coordinate Y (i.e., statement, >= 0)
    Y INT NOT NULL,
    -- Observation type (e.g., 'value' for output, 'input_value' for input etc.)
    TYPE VARCHAR,
    -- Observation value (serialized)
    VALUE VARCHAR,
    -- Raw observation value (unserialized)
    RAWVALUE VARCHAR,
    -- Object type of observation value
    VALUETYPE VARCHAR,
    -- Timestamp of observation
    LASTMODIFIED TIMESTAMP,
    -- Execution time (i.e., for observation values)
    EXECUTIONTIME BIGINT
)

Identifying Arena Executions​

Combination of

  • EXECUTIONID
  • ACTIONID
  • ARENAID

In other words, an instance of the arena test driver with ID ARENAID is executed as part of an action ACTIONID defined by the executed LSL pipeline script EXECUTIONID.

Frequent values for the ARENAID column include

  • execute* - for normal test execution
  • jacoco* - for code measurements
  • pitest* - for mutation testing

Identifying Implementations​

Combination of

  • SYSTEMID
  • ADAPTERID
  • VARIANTID

A code unit with ID SYSTEMID (as stored in the code index) using adapter ADAPTERID. Sometimes, we generate variants of a code unit (e.g., mutated code) which are depicted by VARIANTID.

Identifying Sequence Sheets​

Combination of

  • SHEETID (test name)
  • X (x-coordinate of the cell)
  • Y (y-coordinate of the cell; it identifies the statement)

Examples

  • (X=0,Y) refers to the output value column in a sequence sheet
  • (X=1,Y) refers to the operation column in a sequence sheet
  • (X>1,Y) refers to the input value columns
  • (0,0) refers to the output value by the CREATE statement and stores the instance of the class under test as its output

Note that SHEETID is currently represented as a (string) compound key of

  • test name
  • SYSTEMID

separated by _. For data manipulation, the postfix part needs to be removed:

  • testName_SYSTEMID -> testName

Identifying Observations​

Combination of

  • VALUE
  • TYPE

Frequent values for the TYPE column (i.e., observational types) include

  • value - for output observations part of responses
  • input_value - for input observations part of stimuli
  • op - operation name
  • seq - Java sequence of statement
  • exseq - Test sequence of statements (JUnit like)
  • jacoco_* - for code measurements